Digital Droplet PCR (ddPCR) for Detection and Quantitation of Hepatitis Delt Virus


 Background & Aims

The prevalence of hepatitis delt virus (HDV) far exceeds our expected level, there remains a lack of reliable quantitative assays for HDV RNA detection. We sought to develop a new method based on digital droplet PCR (ddPCR) for HDV RNA quantitative detection.
Methods

With plasmid (pMD19T) containing HDV full-genome, we determined the method for ddPCR-based HDV RNA quantification. To compare various assays for HDV detection, 30 cases diagnosed hepatitis D and 14 controls were examined by ELISA, RT-PCR and ddPCR. 728 HBV-related patients including 182 chronic hepatitis B (CHB), 182 liver cirrhosis (LC), 182 hepatocellular carcinoma (HCC) and 182 liver failure (LF) were screened for HDV infection.
Results

The limit of detection of ddPCR for HDV is significantly low, which lower limit of detection (LLoD) and lower limit of quantitation (LLoQ) to be 5.51 copies/reaction (95% CI: 1.15–6.4*105) and 0.18 copies/reaction (95% CI: 0.0012151- 0.76436), respectively. Among the 44 samples, ELISA detected 30 cases positive for anti-HDV, ddPCR reported 24 samples and RT-PCR reported 10 samples positive for HDV RNA. Moreover, the positive rates of anti-HDV IgG were 1.1%, 3.3%, 2.7% and 7.1% in patients with CHB, LC, HCC, and LF; the detection rates of RT-PCR in HDV RNA were 0%, 16.67%, 15.4% and 20%, however, the detection rates of ddPCR were 0%, 33.33%, 30.77% and 60%.
Conclusion

We establish a high sensitivity and high specificity quantitative HDV RNA detection method based on ddPCR compared to RT-PCR. HBV-related end-stage liver disease, especially liver failure, are associated with a remarkably high rate of HDV infection.


Introduction
Hepatitis delta virus (HDV) was found in the nucleus of hepatocytes in 1977 by Italian scholars and is the smallest known human virus. HDV is a defective RNA virus, and its proliferation and propagation are dependent on the assistance of hepatitis B virus (HBV) to provide the viral envelope (1). Full-length genomic nucleotide sequencing and phylogenetic analyses have identi ed eight genotypes of HDV, with separation among genotypes up to 40% over the full-length sequence. HDV-1, which comprises 4 subgenotypes, is the most prevalent genotype worldwide, and the geographic distributions of genotypes show obvious differences (2). Compared with HBV monoinfection, patients with HDV and HBV coinfection have the most severe form of viral hepatitis. HDV infection signi cantly accelerates disease progression of chronic hepatitis B, which progresses to cirrhosis within 5 years and to hepatocellular carcinoma within 10 years on average (3,4).
The latest meta-analysis showed that 0.16% of the general population is estimated to be positive for anti-HDV antibodies, with approximately 10.6% of patients infected with HBV being coinfected with HDV worldwide (5). Another recent study estimated that 48-60 million individuals are infected globally (6). Additionally, the prevalence of HDV differs signi cantly among geographic regions. One study conducted in 2011 to 2016 in the United States showed that 42% of adult HBsAg carriers have antibodies against HDV (7). In China, a recent meta-analysis showed an anti-HDV antibody prevalence rate of 2.1% among HBsAg carriers and 0.4% among the general population (5). However, the number of hepatitis D cases reported from 2016 to 2020 was 441, 481, 356, 352 and 187 for these years, revealing a large discrepancy from the number based on the prevalence of the national viral hepatitis seroepidemiological survey (8). Approximately 10% of HBsAg-positive individuals in China are estimated to be infected with HDV, and approximately 90 million individuals in China have HBV infection (9,10). Due to the limited ability to effectively detect HDV, infection with this virus is likely to be a major obstacle for achieving the WHO's goal of eliminating viral hepatitis, including chronic hepatitis B, C, and D, worldwide by 2030.
At present, the limitations of HDV laboratory examinations are one of the important reasons why HDV infection is underestimated. All immunocompetent patients infected with HDV can produce anti-HDV antibodies, including IgM and IgG. Positivity for anti-HDV IgM indicates HDV replication, whereas IgG suggests a previous HDV infection and persists for many years. Serum HD antigen is detectable only transiently in blood specimens collected early at the onset of HDV infection, before the rise of antibody titers. In clinical laboratories, ELISA to detect anti-HDV IgM or IgG is the most common measure to screen for HDV infection. On the one hand, there is a lack of uniform quality standards for kits produced by different manufacturers, and the results of different laboratories are  not comparable; on the other hand, test results are inaccurate if the virus infection occurs within the window period and is closely related to the patient's own immune status (11). Although the quantitative microarray antibody capture assay, which has high speci city and sensitivity, is likely an ideal tool for population screening (12), it is not widely used. However, HDV RNA detection is considered the "gold standard" for diagnosing hepatitis D infection. Recently, RT-PCR (real-time reverse-transcriptase polymerase chain reaction) assays have been employed for relatively quantitative detection of HDV, and the sensitivity and accuracy have improved signi cantly (13). The detection limit of RT-PCR for HDV is approximately 1000 copies/µl, but it is not useful for all the various genotypes (1). Moreover, due to the complexity of primers and probe design, this assay is not well standardized, and results from different laboratories are di cult to compare (13).
The droplet digital PCR (ddPCR) system is based on sample dropletization. After PCR ampli cation, the concentration of the target molecule can be quanti ed to 1 copy/µl. In this study, we developed and present a new assay for HDV measurement based on ddPCR that is characterized by improved sensitivity and accuracy compared with RT-PCR. Here, we compare ELISA, RT-PCR and ddPCR for HDV RNA detection and explore the prevalence and quantity of HDV in patients with HBV-related liver disease.

Patients And Methods
Clinical data Enzyme-linked immunosorbent assay (ELISA) Serum anti-HDV IgM and anti-HDV IgG were determined with enzyme-linked immunosorbent assays (ELISAs) using commercial kits (Abbexa Ltd, Cambridge, UK) according to the manufacturer's instructions. Brie y, serum samples were diluted at 1:11 with diluent; the appropriate positive and negative controls were established. Sample and blank wells on the precoated plate were added, and the plate was covered and incubate at 37°C for 30 min. Then, the plate was washed 5 times with wash buffer, the detection reagent was added, and the plate was incubated and washed as above. Finally, 50 µl of TMB substrates A and B were added, followed by incubation at 37°C for 10 min; 50 µl of stop solution was added to each well, and optical density was measured at 450 nm.

RNA extraction and reverse transcription
Total RNA was extracted using a QIAamp Viral RNA Mini Kit (QIAGEN, Valencia, CA) from 150 µl serum samples according to the manufacturer's instructions. Finally, 50 µl nucleic acid extract was obtained and used for reverse transcription.

Results
Sensitivity and dynamic range of the ddPCR assay to detect HDV The sensitivity of the ddPCR method for HDV detection was assessed using the plasmid pMD19T containing the HDV full genome, as shown in the supplemental materials. A 10-fold serial dilution of the plasmid, ranging from 10 6 to 10 0 , was prepared to test the linearity of the ddPCR method using primer and probe sets targeting the HDV common sequence (Fig. 1A). Linear regression analysis showed that this method has excellent linear correlation between the detected value and expected value, with R 2 = 0.9985 (Fig. 1B). In addition, RT-PCR was used to detect the above serially diluted plasmid, with a reported range of 10 3 to 10 6 copies/reaction. When the target concentration was higher than 10 3 copies/reaction, RT-PCR displayed good linear correlation, with R 2 = 0.9995 (Fig. 1C). These results demonstrate that compared with RT-PCR, ddPCR has a wider detection range, with a lower limit of detection, even reaching 1 copy/reaction.
Determining the LLoD and LLoQ of the ddPCR assay for HDV detection Next, we performed probit analysis with a sigmoid curve to determine the lower limit of detection (LLoD) and lower limit of quantitation (LLoQ) of both ddPCR and RT-PCR. First, HDV RNA was extracted from a positive serum sample and then reverse transcribed to cDNA and tested using ddPCR. The sample was serially diluted 10-fold to concentrations from 10 − 2 to 10 3 copies/reaction. Each concentration was tested in 7 replicates, and the LLoQ and LLoD were determined by probit regression with 95% and 50% repeatable probability, respectively. The ddPCR results showed an LLoQ of 5.51 (95%CI: 1.15-6.4*10 5 ) and an LLoD of 0.18 (95%CI: 0.0012151-0.76436) ( Fig. 2A). The same method was used to determine the LLoD and LLoQ of RT-PCR, with results of Probe and primers Because HDV has eight different genotypes, we selected the common sequence of the 8 genotypes as a detection target. The primer sequences used are as follows: Forward: 5′-CTCGGTAATGGCGAATGGGA − 3′; Reverse: 5′-TTCTTTCCTCTTCGGGTCGG − 3′; probe: 5′-FAM-GCTCTCCCTTAGCCATCCGAG-TAMRA-3′.

Reverse transcription polymerase chain reaction (RT-PCR)
The template cDNA, primers and probes described above were used. The 20-µl reaction contained 10 µl 2*mix, 0.5 µl of forward primer and reverse primer each, 0.5 µl of probe, 3 µl of template and 5.5 µl of reaction buffer. Thermal cycling using an ABI7500 Real-Time PCR Detection system was performed at 94°C for 3 min, followed by 35 cycles of 94°C for 30 s and 58°C for 45 s.

Droplet digital PCR (ddPCR)
All ddPCR procedures were conducted following the manufacturer's instructions for Droplet Digital PCR System (TargetingOne® Biotech. Co. Ltd. Beijing, China). Brie y, the TaqMan PCR mixture contained 15 µl of 2x ddPCR Supermix, 1.2 µl of primers, 0.6 µl of probe and 2 µl of template, and deionized water was added to a nal volume of 30 µl. The reaction was converted to droplets with the TargetingOne droplet generator, and then the microdroplet sample was ampli ed with a T100 Thermal Cycler. The program was as follows: 95°C for 10 min (DNA polymerase activation), followed by 40 cycles of 94°C for 30 s and 60°C for 1 min (annealing) and an in nite 4°C hold. Positive and negative controls were included for each test. The ampli cation product was assessed using a chip reader to detect FAM signals; the results were analyzed using the analyze software, and HDV cDNA values are reported as copies/µl.

Statistical analysis
For ddPCR assay characterization, the data were analyzed with speci c software to calculate the concentration of the target. The coe cient of determination was assessed by linear regression analysis using GraphPad Prism 8.00. In addition, LLoD and LLoQ were calculated by probit regression analysis with MedCalc software 19.0.4, and the lowest concentrations of 95% and 50% positive samples were detected.
Continuous variables are presented as mean ± standard deviations (SDs). A p value < 0.05 was considered statistically signi cant.

Page 5/15
384.62 (95%CI: 116.10-1047.0) and 4328.89 (95%CI: 1431.00-167353.0), respectively (Fig. 2B). Obviously, both ddPCR and RT-PCR were reliable when testing high-concentration samples, but ddPCR performed more precisely when detecting samples with concentrations lower than 10 3 copies/reaction. Speci city of the ddPCR assay for detecting HDV To assess the speci city of ddPCR for HDV detection, we used various kinds of plasmids, including plasmids containing the fulllength HBV genome, HCV plasmids, HIV plasmids and HDV plasmids, to perform three independent tests. According to the results, ampli cation of plasmids other than the HDV plasmid led to no positive signal, which con rms the high speci city of the ddPCR method with the primers/probe set targeting HDV (Fig. 3A).
In addition, we evaluated the speci city and effectiveness of ddPCR for HDV detection. We extracted virus nucleic acids from the serum of patients who were clinically con rmed, including cases of monoinfection with HBV, coinfection with HBV and HCV, coinfection with HBV and HIV, and coinfection with HBV and HDV. The above samples were tested using the ddPCR method in parallel, and positive events were only found for patients with coinfection of HBV and HDV (Fig. 3B). In summary, ddPCR for HDV detection is a highly speci c method that signi cantly reduces false positive results.
Detection e ciencies of ddPCR, RT-PCR and ELISA for HDV In this study, 44 samples were collected from clinical patients, and the clinical information is summarized in Table 1 and supplement   Table 1. Thirty samples were from patients clinically diagnosed with hepatitis D, and 14 samples were determined to be HBV infection alone. First, we performed ELISA to detect anti-HDV IgG, and 30 samples were positive, consistent with the clinical diagnosis.
Afterwards, viral nucleic acids were extracted from 44 samples simultaneously, and cDNA products were obtained after reverse transcription. These products were used for ddPCR, RT-PCR at the same time.  The results of nucleic acid analysis showed that RT-PCR was only able to detect 10 positive samples; ddPCR detected 24 positive samples, 14 more than RT-PCR (Fig. 4A, B). Based on our results, positivity for anti-HDV IgG indicates that the patient had a previous infection but that perhaps no virus remained due to the application of antiviral drugs. For the presence of low levels of HDV in patients, the detection e ciency of ddPCR was signi cantly higher than that of RT-PCR.

Detection of HDV in patients with HBV-related diseases by ELISA, RT-PCR and ddPCR
A total of 728 samples were examined in this study, including 182 from patients with chronic hepatitis B, 182 from patients with HBVrelated liver cirrhosis, 182 from patients with HBV-related liver failure and 182 from HBV-related hepatocellular carcinoma (Supplement Fig. 1), which basic information were summarized in supplement Table 2. First, anti-HDV IgG and anti-HDV IgM in these samples were detected using ELISA methods. As shown in Table 2 we established is a high-sensitivity and high-accuracy detection method for HDV, especially for HDV RNA less than 10 2 copies/µl, which is signi cantly superior to RT-PCR. In addition, we found that HDV prevalence differs among patients with chronic hepatitis B, HBV-related liver cirrhosis, HBV-related liver failure and HBV-related hepatocellular carcinoma. Patients with con rmed chronic hepatitis B showed a relatively low HDV infection rate of 1.1% according to anti-HDV IgG compared with those with liver cirrhosis, hepatocellular carcinoma, and liver failure.
In patients with liver cirrhosis and hepatocellular carcinoma, HDV infection rates were 3.3% and 2.7% according to anti-HDV IgG, respectively. In patients with liver failure, anti-HDV IgG positivity was detected in 7.1%, which indicated that coinfection of HDV and HBV signi cantly accelerates the progression of liver disease.

Discussion
It is important to develop an accurate HDV RNA quantitative test, which is critical for HDV diagnosis and guiding treatment response.
Several meta-analyses in recent years have reported an infection rate of HDV of approximately 0.8% of the general population and 13.02% of the HBsAg-positive population (6), even though the anti-HDV positive rate increases by 3-4 times among those with liver diseases (5). According to estimates, there are approximately 12 million people infected with HDV globally. Nevertheless, detection of HDV is often neglected clinically. There may be two reasons: on the one hand, many clinicians lack awareness of the serious disease consequences caused by HDV infection; on the other hand, there are no standard and accurate methods for detecting HDV infection (14,15). There have been many efforts for HDV detection, including HDV antibody screening by ELISA (16-18) and quantitative microarray antibody capture assays (12). However, there is a lack of progress in testing HDV RNA quantitatively, and RT-PCR, which is insu cient regarding sensitivity and accuracy, remains the method for con rming the diagnosis and management of patients (19)(20)(21)(22). Therefore, a method of HDV RNA quantitation with high sensitivity and speci city for the con rmation of HDV infection and treatment monitoring is urgently needed.
In this study, we developed a new method of detecting HDV RNA based on ddPCR, which enables absolute quanti cation of the serum virus with excellent sensitivity and high speci city. ddPCR is a very sensitive and reproducible technique that has been used for testing in various elds. Studies have indicated that ddPCR can be used for the detection and quantitation of HBV cccDNA in the liver of individuals with occult HBV infection (23), for the detection of TP53 deletions and point mutations in chronic lymphocytic leukemia (24), and as a more accurate tool for SARS-CoV-2 detection in low-viral load specimens (25). The more accurate HDV RNA detection method established in this study signi cantly improves the diagnostic ability of HDV infection, including con rmation of HDV infection, especially in patients with low viral loads (cases that are underdetermined by RT-PCR), monitoring the treatment effect of various antiviral drugs and evaluation of disease progression.
ddPCR methods have obvious superiority in HDV RNA detection compared with other available methods. Compared with RT-PCR, our results showed that the sensitivity threshold of these assays was approximately 10 3 copies/µl, which is signi cantly lower than the sensitivity of ddPCR at 1 copy/µl. Moreover, the results obtained from different laboratories are often not comparable due to the use of different primer sets and the nonuniformity of the ampli ed RNA region (26-28). More importantly, commercial and in-house RT-PCR assays in 55% of laboratories often underestimate or fail to quantify HDV viremia, according to the French national quality control study (29). Therefore, considering the limitation of RT-PCR assays for quantitative HDV RNA, the development of more accurate methods for nucleic acids is particularly necessary at the present time. In addition, detecting HDV-speci c IgM or IgG with ELISA is an indispensable approach, especially for the screening of a large range of HBsAg-positive populations. Nonetheless, the widow period for detection is relatively short (30). Anti-HDV IgM typically appears in serum at 2 to 3 weeks after the onset of symptoms and disappears by 2 months after acute HDV infection. Therefore, anti-HDV IgG, as a serologic scar, is commonly used for HDV screening because it can persist in serum after the resolution of acute HDV infection and in chronic HDV infection. Furthermore, anti-HDV detection is usually a false negative in immunode ciency patients, such as those with AIDS. However, con rmation of HDV infection still relies on detectable HDV RNA as a "gold standard" for HDV diagnosis and management.
In addition, we screened HDV infection in different patient cohorts, including cases of chronic hepatitis B, HBV-related cirrhosis, HBVrelated HCC, and HBV-related liver failure. Most striking us is the nding that the HDV infection rate was as high as 7.4% in patients with HBV-related liver failure and that it was relatively low in patients with cirrhosis or HCC and lowest in patients with chronic hepatitis B. Moreover, we developed RT-PCR and ddPCR for all anti-HDV-positive samples; expectedly, the detection rate of HDV RNA with ddPCR was signi cantly higher than that of RT-PCR. Such RNA-undetectable but anti-HDV IgM-positive samples might be due to RNA degradation during storage or cases with no active viral replication.

Declarations
Data availability The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
Ethics approval: All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5). Informed consent was obtained from all patients for being included in the study.

Plant Reproducibility
Not applicable.

Clinical Trials Registration
Not applicable.
Authors' contributions: F.R. and Z.D. designed the experiments. L.X., and X.Z. performed the experiments and wrote the manuscript.
Y.C., Z.F. and Y.T. prepared the samples and collected the data. H.Z. performed statistical analyses. All authors have read and approved the submission of the manuscript.
Con icts of interest: The authors who have taken part in this study declare that they do not have anything to disclose regarding funding or con icts of interest with respect to this manuscript.  Determining the LLoD and LLoQ of ddPCR assay for HDV detection. (A) The probit analysis sigmoid curve was used to determine the lower limit of detection (LLoD) and lower limit of quantitation (LLoQ) of ddPCR. The HDV positive sample with the determined concentration was diluted in a 10-fold series, each concentration repeated 7 times for ddPCR HDV detection. The concentration corresponding 95% probability on curve represent LLoQ and 50% represent LLoD. (B) The probit analysis sigmoid curve was used to determine the lower limit of detection (LLoD) and lower limit of quantitation (LLoQ) of RT-PCR. The analytical method conducted as the same as (A).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. supplementmaterial.docx